technical-runbook

## Purpose Produce a runbook that enables any trained engineer - with no prior knowledge of this specific service - to respond to an alert, diagnose a failure, and execute recovery procedures. **A runbook is not a tutorial or an architecture overview.** It is a rapid-reference guide for someone who is paged at 3 AM and needs to resolve an incident in 15 minutes, not learn the system from scratch. ## Input **Works best with:** The name of the service being documented. **Also valuable:** The alert definitions, common failure modes, diagnostic commands, architecture overview, and escalation contacts. **Example invocation:** `Write a runbook for the PayFlow payment gateway. It's a PHP 8.3 app on Nginx/PHP-FPM, MySQL primary + replica, Redis for sessions/cache, and a queue worker process for async jobs. Alerts fire for: high 5xx rate, slow DB queries, queue depth > 1000, and PHP-FPM pool exhaustion.` ## Key Concepts ### What a Good Runbook Contains (Google SRE Standard) 1. **Overview** - What the service does, its SLA, and its dependencies (30 seconds to read) 2. **Monitoring and Dashboards** - Where to look when something is wrong 3. **Alert Response Procedures** - One procedure per alert: diagnosis, mitigation, escalation 4. **Common Failure Modes** - The top 5-10 failures that happen repeatedly 5. **Diagnostic Commands** - The exact commands to run for each symptom 6. **Escalation Path** - Who to call when the runbook cannot resolve the issue 7. **Maintenance Procedures*...